ECE 5755 - Modern Computer Systems and Architecture

Fall. 3 credits. Letter grades only.

U. Gupta.

This Master’s level course is designed to provide a hardware-centric overview of computer systems used in modern computing platforms. From the bottom up we will study the architecture of processor architectures (e.g., pipelined CPUs, ISA, RISC vs. CISC, out-of-order execution) and memory systems (e.g., memory hierarchy, caching, DRAM memories). We will understand how to evaluate the performance of modern processors and exploit parallelism in applications. This includes parallelization across multi-core CPUs, GPUs, and specialized hardware. Through ands-on assignments and an open-ended project students will develop a holistic understanding of modern computer systems and how they are designed.

Outcome 1: Demonstrate an understanding of micro-architectural principles foundational computer architectures including processor pipelines, caching and memory systems, multi-core designs.

Outcome 2: Demonstrate an understanding of modern computer systems such as software parallelization strategies (e.g., ILP, TLP, DLP), hardware accelerators (e.g., GPUs, FPGAs, ASICs), and modern mobile and datacenter scale systems.

Outcome 3: Demonstrate the ability to analyze performance bottlenecks of software applications running on hardware platforms and optimize systems to maximize performance and efficiency.